Objectives The disappearance of the North American plesiadapoids (stem primates, or plesiadapiforms) in the latest Paleocene has been attributed to competition with rodents over dietary resources. This study compares molar morphology of plesiadapoids and early rodents to assess whether all taxa were adapted to consuming foods of the same structural properties with similar mechanical efficacy. Materials and Methods Micro-CT scans of second mandibular molars (M2s) of plesiadapoids (n = 181) and ischyromyid (early fossil) rodents (n = 13) were evaluated using Dirichlet normal energy (DNE), a dental topographic metric that quantifies the curvature of a tooth's occlusal surface, independent of the orientation of the occlusal plane; this metric can be used to infer diet. Results Comparisons of DNE values for plesiadapoids and rodents show that rodents shared functionally similar dental morphology with at least some plesiadapid plesiadapoids and thus were likely adapted to processing foods with similar physical properties. However, the DNE values for rodents contrast markedly with those for the other two plesiadapoid families, the Carpolestidae and Saxonellidae. Conclusions It is unlikely that direct competition over food resources with rodents played a major role in the extinction of carpolestids and saxonellids, as members of these families were capable of consuming a range of foods that were not accessible to rodents. Although several plesiadapid species overlap with rodents in their range of DNE values, only three overlap in time. One of these (Plesiadapis cookei) may have been too large to be in direct competition with rodents, another (Plesiadapis dubius) has DNE values substantially different (higher) than those of rodents, whereas the third, Chiromyoides, has teeth of both a similar size and DNE value to those of Clarkforkian rodents. If dietary niche overlap with rodents played a direct role in the decline of plesiadapiforms, it can only have potentially done so for Chiromyoides.
- dental topography
- dietary inference